Gamma-ray Large Area Space Telescope (GLAST) Large Area Telescope (LAT) Anticoincidence Detector (ACD) Subsystem Preliminary Design Report

Transcription

1 Document # Date Effective LAT-TD D2 12/06/01 Prepared by Supersedes George Shiblie None GLAST LAT TECHNICAL REPORT Subsystem/Office Anticoincidence Detector Subsystem Document Title LAT Anticoincidence Detector Subsystem Preliminary Design Report Gamma-ray Large Area Space Telescope (GLAST) Large Area Telescope (LAT) Anticoincidence Detector (ACD) Subsystem Preliminary Design Report

2 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 2 of 65 CHANGE HISTORY LOG Revision Effective Date Description of Changes DCN # 1 12/06/01 Initial Release

3 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 3 of 65 Table of Contents 1 Purpose Acronyms and Definitions Definitions Applicable Documents ACD Specifications GLAST References LAT and ACD Supporting Documentation GSFC Procedure and Guidelines Publications ACD Drawings Introduction LAT Science Requirements LAT Technical Description ACD Design Overview ACD Subsystem Organization ACD Subsystem WBS ACD Subsystem Deliverables and Receivables ACD Requirements and Specifications ACD Development and Prototyping ACD Subsystem Design Detector Design Mechanical Design Thermal Design Electronics Design ACD Subsystem Interface Description Mechanical and Thermal Interface Control Electrical Interface Control ACD Safety and Mission Assurance Reliability Parts and Materials Quality Control and Work Order Authorization (WOA) Safety Contamination Control ACD Assembly, Integration and Test Key Mechanical Components Assembly Flow Key Electrical Components Assembly Flow Integration and Test Ground Support Equipment ACD Key Milestones and Schedule Key Level III and Level IV Milestones Schedule...65

4 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 4 of 65 List of Figures Figure 1. View of the LAT Science Instrument with one Tracker tower module and one Calorimeter module pulled away from the Grid. GLAST is a 4 4 array of identical Tracker and Calorimeter modules and the ACD tile cover...13 Figure 2. ACD Subsystem Organization...15 Figure 3. Fraction of events giving a backsplash signal as a function of incident photon energy and ACD threshold. Black dots: results from 1997 SLAC test. Red and blue dots: results from two separate runs at CERN in Figure 4. Three of the scintillator tiles with waveshifting fiber readout produced for the 1999 beam test Figure 5. The ACD for the Beam Test Engineering Model (BTEM). Left: individual wrapped tiles and phototubes mounted on a support structure. Center: one of the electronics cards. Right: the full ACD with its light shield...21 Figure 6. Pulse height spectra from the 13 ACD tiles on the BFEM. In all cases, the signal peak due to charged particles is cleanly separated from the noise...21 Figure 7. The overall layout of the GLAST LAT ACD system, showing the scintillator tile placement...23 Figure 8. Block diagram of the active ACD detector system...23 Figure 9. Scintillator tile with waveshifting fibers in the flight design. Alternate fibers are routed to separate phototubes for redundancy...24 Figure 10. Pulse height distributions of muons from four phototubes...25 Figure 11. Efficiency for flight prototype tiles...25 Figure 12. Technique for sealing gaps on the top surface of the ACD. The ribbons are made of square scintillating fibers...25 Figure 13. ACD Assembly with Exploded View...27 Figure 14. Base Electronics Assembly...28 Figure 15. BEA Corner Mount...28 Figure 16. BEA Center Mount Side View...29 Figure 17. BEA Center Mount Front View...29 Figure 18. Tile Shell Assembly...29 Figure 19. Shell Assembly with Flexures...30 Figure 20. TDA Tie-down Orientation...30 Figure 21. Key Hole Groove...31 Figure 22. Flat and Bent Tile Assembly...31 Figure 23. Clear Fiber Connector...32 Figure 24. PMT Fiber Connector...32 Figure 25. Finite Element Model...33 Figure 26. Corner Boundary Condition...35 Figure 27. Mid-Side Boundary Condition...35 Figure 28. Second Mode Shape...37 Figure 29. First Mode Shape...37 Figure 30. Exterior MLI...39 Figure 31. Towers and Grid...39 Figure 32. ACD Electrical System Block Diagram...41 Figure 33. Analog ASIC Design...42 Figure 34. Power Distribution...44

5 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 5 of 65 Figure 35. Digital ASIC Block Diagram...46 Figure 36. HVBS Block Diagram...46 Figure 37. HVBS Schematic...47 Figure 38. PMT Biasing Circuitry...48 Figure 39. Reliability Allocation and Apportionment...52 Figure 40. Power Supply Reliability Tradeoff Study...53 Figure 41. Typical Corner Joint...57 Figure 42. ACD Subsystem Integration and Test Flow...61

6 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 6 of 65 List of Tables Table 1. ACD Subsystem Work Breakdown Structrure...15 Table 2. ACD Subsystem Level III Key Requirements...18 Table 3. ACD Mass Allocation Table 4. ACD Power Allocation...18 Table 5. Percent Radiation Absorbed...27 Table 6. Material Properties of GLAST components...34 Table 7. Mass Breakdown of ACD Components...34 Table 8. ACD Design Limit Loads (G's)...34 Table 9. Margins of Safety for Corner Fitting...35 Table 10. Margin of Safety of Edge Clips and Flexures...36 Table 11. Deflections under Design Loads...36 Table 12. Modal Frequencies and Effective Masses...37 Table 13. Temperature Predictions...39 Table 14. TSS Optical Properties...40 Table 15. TSS Orbital Parameters...40 Table 16. FMEA 2R Severity Classification Summary...51

7 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 7 of 65 1 PURPOSE This document presents the status of the GLAST LAT ACD subsystem design and planning in support of the January 8-10, 2002 LAT I-PDR. 2 ACRONYMS AND DEFINITIONS ACD ADC AEM ANSI/AIAA ASIC BEA BFEM BTEM CAL CIL CMOS COS-B CTE DAQ DOF EEE EGRET EGSE EMC EMI ESD FEM FM FMEA FREE GAFE GARC The LAT Anti-Coincidence Detector Subsystem Analog-to-Digital Converter ACD Electronics Module American National Standards Institute/Aerospace Institute of Aeronautics and Astronautics Application Specific Integrated Circuits Base Electronics Assembly Balloon Flight Engineering Model Beam Test Engineering Model The LAT Calorimeter Subsystem Critical Items List Complementary Metal Oxide Semiconductor European Gamma-ray Astronomy Satellite Coefficient of Thermal Expansion Data Acquisition Degrees of Freedom Electrical, Electronic, and Electromechanical Energetic Gamma-Ray Experiment Telescope Electrical Ground Support Equipment Electromagnetic Compatibility Electromagnetic Interference Electrostatic Discharge Fine Element Model Flight Module Failure Mode Effect Analysis Front End Electronics GLAST ACD Front End Analog ASIC GLAST ACD Readout Controller Digital ASIC

8 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 8 of 65 GEVS GLAST GOLF GUI HEPA HLD HVBS IC ICD IDT I&T IRD JSC LAT MECO MIP MGSE MLI MOSIS MPLS NPSL PAIP PAPL PCB PDR PHA PMT PPCP PPL P&SA PVM QA RXTE SAM SAS General Environmental Verification Specification Gamma-ray Large Area Space Telescope Global Oscillations at Low Frquencies Graphic User Interface High Efficiency Particle Air High Level Discriminator High Voltage Bias Supply Integrated Circuit Interface Control Document Instrument Development Team Integration and Test Interface Requirements Document Johnson Space Center Large Area Telescope Main Engine Cut-off Minimum Ionizing Particle Mechanical Ground Support Equipment Multi-Layer Insulation Metal Oxide Semiconductor Implementation System Multi-purpose Lift Sling NASA Parts Selection List Performance Assurance Implementation Plan Program Approved Parts List Printed Circuit Board Preliminary Design Review Pulse Height Analysis Photomultiplier Tube Parts Program Control Plan Preferred Parts List Performance and Safety Assurance Performance Verification Matrix Quality Assurance Rossi X-Ray Timing Explorer Safety Assurance Manager Science Analysis Software

9 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 9 of 65 SCL SEL SEU SINDA SLAC S&MA SOHO SRD SRR SSC SSPP TACK TDA T&DF TID TBD TBR TQFP TSA TSS TKR VME WBS WSB WSF WOA Spacecraft Command Language Single Event Latch-up Single Event Upset Systems Improved Numerical Differencing Analyzer Stanford Linear Accelorator Center Safety and Mission Assurance Solar and Heliospheric Observatory Science Requirements Document System Requirements Review Science Support Center System Safety Program Plan Trigger Acknowledge Tile Detector Assembly Trigger and Data Flow Subsystem (LAT) Total Ionizing Dose To Be Determined To Be Resolved Thin Quad Flat Package Tile Shell Assembly Thermal Synthesizer System The LAT Tracker Subsystem Versa Module Eurocard Work Breakdown Structure Wave Shifting Bars Wave Shifting Fibers Work Order Authorization 2.1 Definitions γ µsec, µs A eff Analysis Background Rejection Backsplash Gamma Ray Microsecond, 10-6 second Effective Area A quantitative evaluation of a complete system and /or subsystems by review/analysis of collected data. The ability of the instrument to distinguish gamma rays from charged particles. Secondary particles and photons originating from very high-energy gamma-

10 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 10 of 65 ray showers in the calorimeter giving unwanted ACD signals. Beam Test Test conducted with high energy particle beams cm centimeter Cosmic Ray Ionized atomic particles originating from space and ranging from a single proton up to an iron nucleus and beyond. Dead Time Time during which the instrument does not sense and/or record gamma ray events during normal operations. Demonstration To prove or show, usually without measurement of instrumentation, that the project/product complies with requirements by observation of results. ev Electron Volt Field of View Integral of effective area over solid angle divided by peak effective area. g unit of gravitational acceleration, g = 9.81 m/s 2 Geometric factor Field of View times Effective Area GeV Giga Electron Volts ev Inspection To examine visually or use simple physical measurement techniques to verify conformance to specified requirements. MeV Million Electron Volts, 10 6 ev ph photons s, sec seconds Simulation To examine through model analysis or modeling techniques to verify conformance to specified requirements Testing A measurement to prove or show, usually with precision measurements or instrumentation, that the project/product complies with requirements. Validation Process used to assure the requirement set is complete and consistent, and that each requirement is achievable. Verification Process used to ensure that the selected solutions meet specified requirements and properly integrate with interfacing products.

11 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 11 of 65 3 APPLICABLE DOCUMENTS The documents listed in Section 3.1 are primary to the organization and specification of the LAT ACD subsystem and its interfaces. References in Sections 3.2 through 3.3 provide key references, test specifications, plans and procedures, and other technical documentation relevant to the design of the ACD. Documents that are relevant to the development of the GLAST mission concept and its requirements include the following: 3.1 ACD Specifications 1. LAT-SS-00016, LAT ACD Subsystem Requirements Level III Specification 2. LAT-SS-00352, LAT ACD Electronics Requirements Level IV Specification 3. LAT-SS-00437, LAT ACD Mechanical Requirements Level IV Specification 4. LAT-SS-00448, LAT ACD Electronics Subsystem Specification 5. LAT-SS-00449, LAT ACD Mechanical Subsystem Specification 6. LAT-TD-00438, Light Collection/Optical Performance Tests 7. LAT-TD-00435, LAT ACD Grounding and Shielding Plan 8. LAT-MD , LAT Performance Assurance Implementation Plan (PAIP) 9. LAT-MD , LAT EEE Parts Program Control Plan 10. LAT-SS , LAT Mechanical Parts Plan 11. LAT-MD , LAT System Safety Program Plan (SSPP) 12. LAT-TD-00430, LAT ACD Assembly, Integration, and Test 13. ACD-QA-8001, ACD Quality Plan 14. ACD-RPT-12001, FMEA & CIL 15. ACD-RPT-1250, Limited-Life Items List 16. ACD-REQ-7002, Requirements for ACD MGSE 3.2 GLAST References 17. Response to AO 99-OSS-03. GLAST Large Area Telescope, Flight Investigation: An Astro- Particle Physics Partnership Exploring the High-Energy Universe. Volume 1: Scientific and Technical Plan. Foldouts: A, B, C, D. 18. GSFC 433-SRD-0001, GLAST Science Requirements Document, P.Michelson and N.Gehrels, eds., July 9, LAT-SS-00010, LAT Instrument Performance Specification. 20. GSFC 433-SPEC-001, GLAST Project Mission System Specification, April 24, GSFC 433-IRD-0001, GLAST Science Instrument Spacecraft Interface Requirements Document, Draft July 14, GSFC 433-MAR-0001, Mission Assurance Requirements (MAR) for Gamma-Ray Large Area Telescope (GLAST) Large Area Telescope (LAT), June 9, GSFC 433-RQMT-0005, GLAST EMI Requirements Document. 24. GSFC 433-OPS-0001, GLAST Operations Concept, Sept 7, LAT-SS-00047, LAT Mechanical Performance Specification. 26. LAT-MD-00099, LAT EEE Parts Program Control Plan, March LAT-MD-00039, LAT Performance Assurance Implementation Plan

12 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 12 of LAT-MD-00033, LAT Work Breakdown Structure, May 9, LAT-TD-00125, LAT Mass and Power Allocation Recommendations 30. Gamma Ray Large Area Space Telescope Instrument Technology Development Program, NRA , NASA Office of Space Science, January 16, LAT and ACD Supporting Documentation 31. LAT-SS-00363, LAT Dataflow Subsystem Specification ACD-TEM Interface 32. LAT-DS-00241, LAT Mechanical Systems ACD to Grid Interface Control Description 3.4 GSFC Procedure and Guidelines 33. GPG , Preparation and Handling of Alerts and Safe Alerts 34. GPG , Product Processing, Inspection and Test 35. GPG , Mishap, Incident, and Close Call Investigation PG , Goddard Non-Conformance Reporting and Corrective Action System Configuration Control Board (CCB) 3.5 Publications 37. W.B.Atwood et al., NIM A446 (2000), do Couto e Silva et al., NIM A474/1 (2001), ACD Drawings 39. GE , ACD Assembly Drawing

13 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 13 of 65 4 INTRODUCTION G LAST is a next generation highenergy gamma-ray observatory designed for making observations of celestial gamma-ray sources in the energy band extending from 20 MeV to more than 300 GeV. It follows in the footsteps of the Compton Gamma Ray Observatory EGRET experiment, which was operational between The GLAST Mission is part of NASA's Office of Space and Science Strategic Plan, with launch anticipated in The principal instrument of the GLAST mission is the Large Area Telescope (LAT) that is being developed jointly by NASA and the US Dept. of Energy (DOE) and is supported by an international collaboration of 26 institutions led by Stanford University. The GLAST LAT is a high-energy pair conversion telescope that has been under development for over 7 years with support from NASA, DOE and international partners. It consists of a precision converter-tracker, CsI hodoscopic calorimeter, plastic scintillator anticoincidence system and a data acquisition system. The design is modular with a 4 4 array of identical tracker and calorimeter modules. The modules are ~ 38 x 38 cm. Figure 1 shows the LAT instrument concept. 4.1 LAT Science Requirements The GLAST science requirements are given in Reference 18. An updated set of requirements, as they pertain to the LAT science instrument, is specified in Reference 19. General constraints and requirements on the instrument design are specified in GLAST mission documents (References 20, 21 and 24). The flowdown of the science requirements and instrument constraints to the LAT design is summarized in Foldout-D of our NASA proposal (Reference 17). 4.2 LAT Technical Description Figure 1. View of the LAT Science Instrument with one Tracker tower module and one Calorimeter module pulled away from the Grid. GLAST is a 4 4 array of identical Tracker and Calorimeter modules and the ACD tile cover. The LAT science instrument consists of an Anti Coincidence Detector (ACD), a silicon-strip detector Tracker (TKR), a hodoscopic CsI Calorimeter (CAL), and a Trigger and Data Flow system (T&DF). The principal purpose of the LAT is to measure the incidence direction, energy and time of cosmic gamma rays while rejecting background from charged cosmic rays and atmospheric albedo gamma rays and particles. The data, filtered by onboard software triggers, are streamed to the spacecraft for data storage and subsequent transmittal to ground-based analysis centers. The Tracker provides the principal trigger for the LAT, converts the gamma rays into electron-positron pairs, and measures

14 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 14 of 65 the direction of the incident gamma ray from the charged-particle tracks. It is crucial in the first levels of background rejection for providing track information to extrapolate cosmic-ray tracks to the ACD scintillator tiles, and it is important for further levels of background analysis due to its capability to provide highly detailed track patterns in each event. 4.3 ACD Design Overview The primary task of the GLAST ACD is to detect energetic cosmic ray electrons and nuclei for the purpose of removing these backgrounds. It is the principal source for detection of other than gammaray particles. This detector element covers the Tracker. It consists of an array of 89 plastic scintillator tiles (1 cm thick, various sizes), plus eight (8) scintillating fiber "ribbons" that cover the gaps between the tiles. Signals produced by the ACD are used by the T&DF system to identify cosmic ray electrons and nuclei entering the instrument. 5 ACD SUBSYSTEM ORGANIZATION NASA Goddard Space Flight Center (NASA/ GSFC) has the overall responsibility for the ACD Subsystem by direction of Peter F. Michelson, the Instrument Principal Investigator (IPI). GSFC s responsibility to NASA is identified, with management oversight and concurrence from P.F. Michelson. David Thompson of GSFC, ACD Subsystem Manager, has overall responsibility for the ACD Subsystem of the GLAST LAT instrument. The ACD Instrument Manager, Thomas Johnson, performs the project management. The ACD design and development is done at the NASA Goddard Space Flight Center (NASA/ GSFC). Figure 2 shows the organization of the ACD program.

15 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 15 of 65 ACD Subsystem Dave Thompson, Subsystem Manager ACD management Tom Johnson, Manager Cristina Doria-Warner - Financial Resources Dennis Wicks - Scheduling ACD Design and Science Alexander Moiseev,Lead Jonathan Ormes, Robert Hartman Jay Norris ACD System Engineering George Shiblie Mike Amato ACD Simulations Heather Kelly Taro Kotani Alexander Moiseev ACD Reliability and Quality Assurance Patricia. Huber, Tavi Alvarez, Quality Tony DiVenti, Reliability Nick Virmani,Thom Perry, Parts Fred Gross, PilarJoy, Materials Dave Bogart,Jim Anderson, Safety Randy Hedgeland, Contamination Tile Shell Assembly Ken Segal, Lead Micrometeroid Shield / Thermal Blanket Ken Segal, Lead Carlton Peters, Thermal Lead Hardware Integration & Test Jim La, Lead Alexander Moiseev Bob Hartman Mission Integration & Test Support Bob Hartman, Lead Tile Detector Assemblies A. Moiseev, Lead Base Electronics Assembly Glenn Unger, Lead D.Sheppard, S. Singh, R. Baker, A. Ruitberg Flight Software moved to LAT Flight Electronics LAT Insrument Integration & Test Support Jim La, Lead Ground Support Facilities & Equipment B Jim La, Lead Glenn Unger Ken Segal Figure 2. ACD Subsystem Organization 6 ACD SUBSYSTEM WBS Table 1 provides a top-level summary of the ACD WBS to Level 5. Table 1. ACD Subsystem Work Breakdown Structrure WBS Task ACD ACD Management Project Management System/Subsystem Engineering Science Support Simulations

16 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 16 of 65 WBS Task Reliability and Quality Assurance Reliability Safety Flight Assurance Parts Control Materials Tile Shell Assembly (TSA) Tile Shell Assembly (TSA) Analysis/Design Tile Detector Assemblies (TDAs) Shell Subassembly Tile Mounting Hardware & Assemblies Reserved Reserved Cal Unit components I&T Flight Model TSA I&T Base Electronics Assembly (BEA) Base Electronics Assembly (BEA) Analysis/Design BEA Base Frame Electronics Assembly High Voltage Power Supply Analysis/Procurement Analog ASIC Design/Procurement Digital ASIC Design/Procurement Front-Ent Electronics and Event (FREE) Circuit Card Assembly (CCA) Reserved Reserved BEA Electronics Mounting/Assembly Hardware A Reserved B Reserved C Cal Unit BEA I&T D Flight Model BEA I&T E Photo-Multiplier Tubes (PMT) & Divider Assembly Micrometeoroid Shield/Thermal Blanket Assembly Micrometeoroid Shield/Thermal Blanket Assembly Micrometeoroid Shield/Thermal Blanket Assembly Mounting/Assembly H/W Reserved Reserved Reserved Micrometeoroid Shield/Thermal Blanket Assembly Flight Unit I&T Reserved Hardware/ Software Integration & Test Miscellaneous LAT Interface/Assembly Hardware Reserved

17 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 17 of 65 WBS Task Reserved Calibration Model Hardware/Software I&T Flight Model Hardware/Software I&T LAT Instrument Integration & Test Reserved Reserved Cal Unit I&T w/lat (Support) Flight Unit I&T w/lat (Support) Mission Integration and Test Support LAT Flight Unit I&T w/spacecraft (Support) Reserved A Reserved B Ground Support Facilities & Equipment B.1 Reserved B.2 Mechanical Ground Support Equipment B.3 Electrical Ground Support Equipment B.4 Front End Electronics and Event (FREE) Circuit Card Assembly (CCA) Test Set 7 ACD SUBSYSTEM DELIVERABLES AND RECEIVABLES The ACD Subsystem deliverables and receivables due dates are documented in the ACD schedule. The ACD subsystem will deliver the following items to the project: 1. One (1) ACD Subsystem FLIGHT Unit 2. One (1) DAQ I/F test unit (test screening board with digital ASIC) delivered to T&DF for ACD Electronics Module (AEM) Development 3. The parts necessary to assemble 12 Front End Electronics circuit cards (to be assembled at SLAC) delivered to T&DF for AEM Development. 4. Calibration unit components. 5. ACD mechanical and thermal finite-element models. 6. Mechanical ground support equipment for module handling and installation. 7. Design documentation and as-built documentation for the ACD, including the fabrication database. 8. The ACD operating and handling manual, including test scripts. The ACD subsystem will receive the following items for the development of the project: 1. Five ACD Electronic Modules (AEM) units with cable interfaces to ACD. 2. Three (3) Front End Workstations with SCL control executive and LabVIEW GUI. 3. AEM operating and handling manuals. 4. AEM timing diagrams that describes every mode of operation.

18 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 18 of 65 8 ACD REQUIREMENTS AND SPECIFICATIONS The ACD Subsystem requirements and specifications are found in the ACD Subsystem Level III Specification (Reference 1 in Section 3.1). It specifies the ACD requirements necessary to meet the overall LAT system performance (Reference 19 in Section 3.2). The ACD Level III Specification was formally reviewed and approved in April, 2001; however, extensive changes have occurred since then because of descoping to ACD. The change version is under review for final approval. Table 2 is a summary of the most important level III requirements. The ACD Level IV requirements (Reference 2 and 3) are derived for the specific implementation of the LAT ACD and its interface requirements to other LAT subsystems. In this implementation, the major components of the ACD subsystem and their specifications are: ACD Electronics Subsystem Specification, LAT- SS-00448, Reference 4 ACD Mechanical Subsystem Specification, LAT- SS-00449, Reference 5 The resources allocated to the ACD subsystem are documented in Reference 29 and are summarized below. As shown in Table 3, the baseline ACD mass was established at the System Requirements Review at kg. Based on ANSI/AIAA standards a reserve of 60.9 kg was identified, but only 5.7 kg of that reserve have been allocated to the ACD. Thus, the ACD is designing to a mass limit of 205 kg. Exceeding this limit requires configuration control board allocation from the mass reserve. The LAT IDT is currently reviewing a formal request to change the allocated mass from to kg with no change to the reserve. Table 4 shows a similar summary for the conditioned power allocated to the ACD. The ACD is allocated 31 watts of conditioned power. Table 2. ACD Subsystem Level III Key Requirements Parameter Requirement Detection of Charged Particles Fast VETO Signal PHA signal False backsplash VETO rate High Threshold (Heavy Nuclei) Detection Size Mass Power Instrument Lifetime >0.3 MIP nominally average detection efficiency over entire area of ACD (excluding bottom row of tiles) Logic signal nsec after passage of charge paricle For each phototube, pulse height measurement for each Trigger Acknowledge (TACK) <10MIP, precision of <0.02 MIP or 5% >10MIP, precision of <1 MIP or 2 % < 20% false VETOs due to calorimeter backsplash at 300 GeV > 25MIP detection of highly-ionized particles. Outside: 1796 x1796 x 1015 mm Inside Grid: 1574 x 1574 x mm Inside TKR: x x 650 mm < 205 kg < 31 Watts (conditioned) Minimum 5 yrs Table 3. ACD Mass Allocation. Mass (kg) SRR Est. (Adj.) ANSI/ AIAA Reserve Recom'd Subsystem Reserve Allocation Subsystem Budget Allocation Table 4. ACD Power Allocation Power (Watts) SRR Est. (Adj.) 24.1 ANSI/ AIAA Reserve Recom'd Subsystem Reserve Allocation Subsystem Budget Allocation

19 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 19 of 65 9 ACD DEVELOPMENT AND PROTOTYPING Plastic scintillator has been used in all previous gamma-ray telescopes for anticoincidence against charged particles; however, segmenting the tiles is a new design. The GLAST LAT design is the first instrument to have the twin requirements of high efficiency for charged particles and low susceptibility to backsplash self-veto. Segmenting the scintillator attacks the backsplash problem but increases the challenge of achieving high efficiency. Unlike a monolithic anticoincidence detector, which has no gaps and uses many phototubes to collect light, a segmented anticoincidence detector may have gaps between segments and can use few phototubes to collect light. A series of development and trade studies have been carried out in order to verify and optimize the ACD design: A trade study of readout techniques led to the conclusion that conventional PMT tubes remained the light collector of choice < A study was conducted to determine the most effective way to collect the light from the scintillator tiles. Wave-shifting bars (WSB) and direct-coupled PMTs were considered as alternatives to the baselined wave-shifting fibers (WSF). Laboratory tests were conducted to compare their light collection responses. The test results conclude that the WSF provide much better uniformity of response and that it is still the best choice. This result was reported in the report on the NASA Technology Development contract. For the 1997 beam test at SLAC, we built a set of 15 scintillator tiles with waveshifting fiber/photomultiplier readout. These tiles demonstrated the required efficiency and quantified the backsplash generated in the calorimeter. Results were reported in the beam test paper, W.B.Atwood et al., NIM A446 (2000), 444 A follow-up test of backsplash using the same scintillator tiles was carried out at CERN in 1999, in order to extend the measurement to 300 GeV. The probability of a backsplash signal (seen in Figure 3) is given by the following empirical formula: A 55 P = E backsplash + E 144 x 10 thr + W here E is the energy of incident electron/photon in GeV E thr is the threshold value in units of mip X is the distance from the top of calorimeter A is area in cm 2 P backsplash is the probability that there was an energy deposition above Ethr in 1cm scintillator

20 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 20 of 65 Figure 3. Fraction of events giving a backsplash signal as a function of incident photon energy and ACD threshold. Black dots: results from 1997 SLAC test. Red and blue dots: results from two separate runs at CERN in 1999 Also in 1999, we built a set of 13 scintillators with waveshifting fiber/photomultiplier readout for a SLAC beam test of a prototype tower. Several different tile/fiber configurations were made, with examples shown in Figure 4. The assembled prototype tower ACD is shown in Figure 5. Figure 4. Three of the scintillator tiles with waveshifting fiber readout produced for the 1999 beam test.

21 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 21 of 65 Figure 5. The ACD for the Beam Test Engineering Model (BTEM). Left: individual wrapped tiles and phototubes mounted on a support structure. Center: one of the electronics cards. Right: the full ACD with its light shield. Results on the ACD from the 1999 beam test were consistent with those found in the previous beam test (do Couto e Silva et al., NIM A474/1 (2001), 19). The same ACD was used for the Balloon Flight Engineering Model (BFEM), where it performed successfully on the August, 2001, balloon flight (Figure 6). Data from that flight are still being analyzed. Several studies have been carried out on the segmentation of the LAT ACD. The effect of backsplash can be reduced with a larger number of smaller tiles, but at a cost of increased complexity and expense. A detailed study, based on simulations, < cdpdr/acd_segmentation.pdf> found that the segmentation could be reduced from the 145 tiles of the AO response to 105 tiles, with minimal loss of performance. Figure 6. Pulse height spectra from the 13 ACD tiles on the BFEM. In all cases, the signal peak due to charged particles is cleanly separated from the noise The effect of gaps between tiles (hermeticity) was studied in detail with simulations. Simple butt joints would require at least 2 mm spacing to allow for thermal expansion, and this spacing is too large to achieve the required overall efficiency of Overlapping the tiles in two dimensions is quite complex to design, build, and test. A one-

22 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 22 of 65 dimensional overlap still leaves some gaps, but these could be covered with ribbons made of scintillating fibers. A detailed study < found this last option to be attractive. The scintillator tile design was carefully studied for the maximization of the light collection. The parameters studied are the following: fiber spacing effect effect of wrapping material - light reflection aluminization of the fiber ends fiber cladding scintillator manufacturer other different designs We measured the absolute efficiency of a tile of the chosen design and estimate the light yield to be around 35 photoelectrons for two PMT operations, and 19 photoelectrons for single PMT operation. Assuming 15% light loss in the clear fiber extension, the efficiency for single PMT operation will be slightly below the required at the nominal threshold of 0.3 MIP. The default-operating mode must be with both PMT's operating, which provides the required efficiency with margin. We also studied the effect of broken fibers and found that we have good redundancy in the fibers. The minimal requirement for launch is that there must be no tile with more than 4 broken fibers, and no more than 3 tiles in total with broken fibers. To be qualified for assembly into the ACD, a tile may have NO broken fibers. Our 4 years' experience working with these detectors provide confidence that, with proper handling of the tiles, there should not be any occurrence of broken fibers. Results from these studies are documented in LAT-TD-00438, Light Collection/Optical Performance Tests. Wherever possible, these design and trade studies were backed up with measurements of test articles, either in the laboratory or in one of the LAT tower prototypes. In particular, More than 45 different tile prototypes with WSF readout have been fabricated by us since Also, a fiber ribbon prototype was built and tested demonstrating at least 6 photoelectrons as required. The result of this work is the design described in the following section.

23 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 23 of ACD SUBSYSTEM DESIGN 10.1 Detector Design. An overview of the ACD detector assembly is shown in Figure 7, and a block diagram of the readout is shown in Figure 8. The following paragraphs will describe the specifics of the detector design, starting with the overall configuration and then working through the specific detector elements Segmentation The ACD segmentation was optimized by evaluating the effective area and geometric factor degradation for high-energy (300 GeV) photons. Taking into account the fact that the first tracker tray with the radiator is situated at >18cm from the calorimeter, we considered those events which enter the ACD side at least 15 cm from the ACD bottom (calorimeter top). With this approach we were able to reduce the number of tiles from 145 to 89, with only a few percent loss in effective area. The final segmentation of active area on the sides is 3 rows of 5 tiles each, plus a bottom row consisting of a single tile. The reduced segmentation design for the ACD is 5 by 5 tiles on the top and 16 tiles on each side (89 tiles in total). Figure 7. The overall layout of the GLAST LAT ACD system, showing the scintillator tile placement. Figure 8. Block diagram of the active ACD detector system.

24 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 24 of Scintillator Tile Design The chosen design is (example shown in Figure 9): 1 cm thick plastic scintillator (BC-408 or ElJen 200) 5 mm spaced, 1.6 mm deep straight grooves 1 mm diameter BCF-91A/MC waveshifting multiclad fibers (WSF) glued into grooves by BC-600 optical cement High light reflecting TETRATEC wrapping Aluminized fiber ends Clear 1.2 mm diameter fibers BCF-98, connected to WSF near the tile, to bring the light to the PMT with minimal light loss (for the longest bundles, the use of waveshifting fibers for the entire length would result in a 30-40% light loss due to absorption; the connection to the clear fibers has a 15% light loss, but the clear fibers themselves add essentially no additional light loss). Figure 9. Scintillator tile with waveshifting fibers in the flight design. Alternate fibers are routed to separate phototubes for redundancy. Performance of two tiles (four phototubes) with this design was verified in the laboratory using atmospheric muons. Figure 10 shows pulse height distributions of muon signals from four phototubes attached to waveshifting fiber bundles from two scintillator tiles. The characteristic Landau distribution is clearly seen, along with the fact that few signals fall below the lower end of the distribution. Figure 11 shows detection efficiency as a function of threshold for the individual phototubes and combinations. The required efficiency can be met with a threshold of 0.3 MIP (minimum ionizing particle). If required, higher efficiency could be achieved with a lower threshold or by combining signals from the two tubes.

25 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 25 of 65 Figure 10. Pulse height distributions of muons from four phototubes Figure 11. Efficiency for flight prototype tiles Hermeticity In order to minimize the effects of gaps between scintillator tiles, all tiles are overlapped by 1cm in one direction, and have gaps in the other. The gaps are required to be ~ 3 mm (0.7mm for wrapping and 2mm for thermal expansion). These gaps are covered by scintillating fiber ribbons, as shown in Figure 12. Simulations show that the ACD as a whole can meet the efficiency requirement with this design Photomultiplier Tube Fiber ribbon covering gap Figure 12. Technique for sealing gaps on the top surface of the ACD. The ribbons are made of square scintillating fibers. Because the ACD requires 194 phototubes, the tube must be small, in addition to having gain and noise characteristics suitable for detecting small numbers of photoelectrons. The tube selected is the Hamamatsu R4443, a ruggedized version of the R647 that we used for the 1997 beam test. The R647 tube (the non-ruggedized version) was used on the RXTE satellite mission and a nearly equivalent R4444 was used successfully on the GOLF instrument aboard the SOHO satellite.

26 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 26 of Thermal Blanket/Micrometeoroid Shield/ACD "Crown" As the outer detector element of the LAT, the ACD must be covered by some sort of thermal insulation. A second requirement in addition to thermal control is protection of the light-sensitive scintillator from a penetration by space debris or a micrometeoroid that could allow a light path. In full sun, even a tiny light leak would disable the TDA. The design issue is that the outer shield must also have very low mass, because nuclear interactions in this inert material will produce gamma rays from neutral pion decay. The experience with the COS-B gamma-ray telescope illustrates that such locally-generated background is not a hypothetical concern: the cosmic-ray-induced gamma-ray background in COS-B exceeded the flux from the high-latitude diffuse gamma radiation. The low mass requirement clearly conflicts with the protection goal. The three design rules that were used to achieve a low background for EGRET were: 1. Make the blanket/shield as lightweight as possible (low mass makes a smaller target). 2. Keep the blanket as close to the scintillator as possible (maximizing the chance that charged particles hitting the blanket will also produce a VETO signal in the scintillator). 3. Minimize the linear path length through the blanket (a shorter path reduces the chance of an interaction). The first two of these design principles are achieved in the manufacture and placement of the thermal blanket/micrometeoroid shield itself, as described in section The third affects the ACD scintillator design. Because the GLAST LAT ACD is box-like, it presents a long straight path through the thermal blanket on the upper (forward) face of the ACD. Simulations show that the flux of gamma-rays induced in this large, flat surface could be significant compared to the diffuse gamma radiation the LAT is trying to measure. Details are given in "Does ACD Need a Crown?" Our solution is to extend the upper side scintillator panels upward around the edge of the top panels. This "crown" of scintillator will flag those cosmic rays that go through the top blanket horizontally. Even then, special analysis will be required, because the ACD signal will not come from the tile that the gamma-ray points toward Mechanical Design The ACD Mechanical Subsystem is comprised of two primary assemblies, the Base Electronics Assembly (BEA) and the Tile Shell Assembly (TSA). The BEA is the mechanical support structure for the ACD. It supports the TSA and houses the ACD electronics as well as provides the mechanical and electrical interfaces to the LAT. It is an aluminum structure and it is mechanically symmetric. The TSA supports the Tile Detector Assemblies (TDA s) and their associated fiber cables as well as the Micrometeoroid Shield/Thermal Blanket. Figure 13 shows an exploded view of the ACD. The ACD Assembly drawing number is (GE ).

27 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 27 of 65 The estimated mass of the ACD is kg and the calculated center of gravity is 0,0,332 mm (in the LAT coordinate system). The 24 electrical cables from the LAT are mated to the ACD using electrical bulkhead connectors on the BEA. 850 mm 1720 mm The calculated average percent gamma radiation absorbed by the ACD is 4.5%, which is less than the required maximum of 6.0%. Table 5 shows a breakdown of the radiation absorbed by each layer of material used in the construction of the ACD. The primary advantages to the ACD mechanical design is that it provides a clean and simple interface to the LAT, is easy to integrate to the LAT, and allows the TSA and BEA to be integrated and tested in parallel. Tile Shell Assembly (TSA) Base Electronics Assembly (BEA) LAT Grid (Trackers not shown for clarity) Figure 13. ACD Assembly with Exploded View Table 5. Percent Radiation Absorbed Layer Equivalent Thickness (cm) Radiation Length (cm) % Radiation Absorbed Nextel Woven Fabric 312* % Solimide Foam* (2.8 cm) % Thermal Blanket % Kevlar* % TDA Wrap (Tetratec) % TDA Wrap (Tedlar)* % Scintillator % GrEP Facesheets % Korex Core* % * Materials radiation length calculated and/or TOTAL 4.46% Base Electronics Assembly The BEA serves as the primary support structure for the ACD. It supports the TSA as well as provides the mechanical interface to the LAT. It is an aluminum structure, which utilizes similar parts in its construction to keep the part count to a minimum. The BEA also houses and protects the ACD onboard electronics as well as provides the electrical interface to the LAT. The BEA is shown in Figure 14.

28 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 28 of 65 Channel (4 places) Machined Aluminum Corner Fitting (4 places) Machined Aluminum +Y +X Connector Bulkhead (24 places) FREE Circuit Card (single row) FREE Circuit Card (double row) FREE Circuit Card Backplane Figure 14. Base Electronics Assembly Closeout The mechanical subsystem in the BEA is the Base Frame. There are four components that make up the Base Frame: corner fittings (4 places), channel sections (4 places), event board backplanes (8 places), and the closeout panels (8 places). On the Y sides there are two single row Event Board Modules. Each Event Board Module has the capability to hold 18 PMT s. Therefore on each of the Y sides there are a total of 36 PMT s. Thirty- Two of the PMT s are for the TDA s and the other 4 PMT s are for the fiber ribbons. We are currently modeling the fiber routing on the Y sides. On the X sides there are two double row Event Board Modules. This is because the output from the TDA s on the top (+Z side) is read out on the X sides. Essentially half of the TDA outputs from the top go to the +X side and the other half go to the X side. Therefore each of the Event Board Modules on the X sides will have inputs from both the top (+Z) and their respective X side. The fiber routing on the X sides has been modeled, and while it is fairly complicated, it is feasible. The Base Frame also provides the mechanical interface between the ACD and LAT. There are eight mounting points between the ACD and LAT, at the four corners and at the center of each side. Figure 15 shows the corner mount design, Figure 16 show the center mount design side view, and Figure 17 shows the center mount design front view. Figure 15. BEA Corner Mount

29 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 29 of 65 Grid Wall Threaded Boss Grid Washer Jam Nut Figure 16. BEA Center Mount Side View Figure 17. BEA Center Mount Front View Tile Shell Assembly The TSA consists of the following components: a composite shell, 8 titanium flexures which attach the TSA to the BEA, 89 TDA s and their associated fiber cables and cable tie-downs, 8 fiber ribbons, and TDA tie-downs. Fiber Ribbons Bent TDA s There are 25 TDA s on the top and 16 TDA s on each side, shown in Figure 13. On the topside there are two different types of TDA s, flat TDA s and bent TDA s, shown in Figure 18, and Figure 22. There are 15 identical flat TDA s and Figure 18. Tile Shell Assembly 10 identical bent TDA s. The bent TDA s are positioned along two edges. They are required to provide routing access for the TDA fibers from the top TDA s to the PMT s, that are mounted on the BEA. The 16 side TDA s are arranged in four rows with five TDA s in each of the upper three rows and one single TDA for the lowest row. The TDA s are shingled in one dimension to eliminate gaps in that dimension; the gap in the other dimension is covered using fiber ribbons. The fiber ribbons are made of scintillating material with PMT connectors on both ends. There are a total of eight ribbons, four ribbons to cover the gaps along the X-axis and the other 4 ribbons to cover the gaps along the Y-axis.

30 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 30 of SHELL ASSEMBLY The composite shell (Figure 19) is constructed using honeycomb panels bonded together at the edges using edge clips. The honeycomb panel facesheet material is 0.5 mm thick M46J/RS-3 and the core is Korex 3/16-2.0psf. The top panel is 51.8 mm thick and the side panels are 26.4 mm thick. 50 mm x 0.5 mm doublers are used around the base of the sides for additional strength. The edges of the panels are filled with lightweight filler and machined so that the panels can be bonded together at their edges as well. Since the edges of the panels are filled and bonded, the facesheets are perforated on the outside to provide venting of the core. Top Edge Clips (4 outer & 4 inner) Side Panel (4 Places) Doubler (8 Places) Corner Flexure (4 Places) Top Panel Middle Flexure (4 Places) Side Edge Clips (4 outer & 4 inner) Figure 19. Shell Assembly with Flexures TITANIUM FIXURES The 8 titanium flexures are all single blade flexures and they are oriented so that they are normal to the center of gravity of the ACD. The flexures are required because of the coefficient of thermal expansion (CTE) difference between the composite shell and the base frame. There are two different types of flexures; one type is used at the corners and the other type is used at the midpoint of the sides. They both have a top and bottom flange and are bolted to the composite shell and base frame. The overall height of the flexures is 75mm and the flanges are 8 mm thick TDA TIE DOWNS The TDA tie-downs (Figure 20) provide the interface between the composite shell and the TDA s. They are kinematic mounts to provide compliance for the thermal expansion and contraction of the TDA s. To minimize weight and radiation absorption the TDA Tiedowns are surface bonded to the composite shell. A stud protruding from each TDA Fixed-Fixed Fixed-Free Fixed-Free Free-Free Nut, Belleville washer, flat washer, Viton gasket (both sides of TDA) Figure 20. TDA Tie-down Orientation

31 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 31 of 65 Tie-down and is used to secure the TDA. This will be accomplished using a combination of seals and gaskets TILE DETECTOR ASSEMBLY There are a total of 89 TDA s with 14 different tile geometries. All tiles are /-0.15 mm thick and there are two basic types of tiles, flat and bent. One (1) mm diameter wave shifting fibers are bonded into the tiles at a spacing of 5 mm. Every TDA will have a connector to transmit the light to the PMT s. In order to facilitate the bonding of the fibers into the bent tiles a keyhole groove (Figure 21) is cut into the tiles before bending. This groove holds the fibers in place while they are bonded in position. The minimum bend radius for the tiles is 4x the thickness of the tile and the minimum bend radius of the fibers is 25x the diameter of the fiber. See ACD-MPML for a list of the materials used for the TDA s. In Figure 21. Key Hole Groove addition, clear fiber cables (Figure 22) are required to transmit the light from the TDA s to the PMT s, except on the bottom two rows. These clear fiber cables are required to facilitate the integration and testing of the ACD, lessen the chance that the TDA s will be damaged during handling, and to minimize light loss over the length of fiber (the attenuation length of clear fiber is over 4x greater than for the wave shifting fiber). To transmit the light from the TDA to the PMT two connectors are required. One connector is required to switch from the wave shifting fiber to clear fiber, and the other connector is required to feed the light into the PMT. Clear Fibers PMT Connector Wave Shifting Fibers Wave Shifting Fibers mirrored on this end Bent Tile TDA Figure 22. Flat and Bent Tile Assembly Flat Tile TDA Wave Shifting/ Clear Fiber Connector C t Clear Fiber Connector The clear fiber connector, Figure 23, will hold up to 68 fibers in two rows. The clear fiber diameter will be 1.2 mm to provide an additional 0.2 mm of alignment tolerance. The connector will be bolted and pinned together to maintain alignment during all environmental conditions. A vent path is also built into the connector to enable venting of the TDA and clear fiber cable PMT Fiber Connector

32 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 32 of 65 The PMT fiber connector, Figure 24, is a circular connector that connects the fiber bundle to the PMT window. It is spring loaded to eliminate overstressing the PMT window, eliminate gapping, and to reduce the positioning accuracy requirement. Interface Plane Clear Fiber Connector Clear Fiber Bundle PMT Coupling Termination Fitting Spring Optical RTV Gasket Vent Path Wave Shifting Fiber Connector Figure 23. Clear Fiber Connector PMT PMT Housing Dynode Module Figure 24. PMT Fiber Connector Micrometeoroid Shield/Thermal Blanket The sole purpose of the micrometeoroid shield is to prevent light leaks to the TDA s caused by the impact of orbital debris. However, it also has the benefit of providing thermal protection as well. A preliminary design of the MS/TB is shown below in Table below. NASA s Micrometeoroid experts at the JSC have been tasked with verifying and optimizing the current design. They have performed orbit simulations using a 3D model of the GLAST Observatory. Several high velocity impact tests have been performed to date and materials studies and characterization is underway. When the final shield design is complete, it will be tested for its thermal properties and then the required amount of thermal blanketing will be added to the micrometeoroid shield. Layer Thickness (cm) Areal Density (g/cm 2 ) Kevlar (innermost layer) Solimide Foam Nextel Woven Fabric Solimide Foam Nextel Woven Fabric Solimide Foam Nextel Woven Fabric Solimide Foam Nextel Woven Fabric Thermal Blanket (outer layer) TOTAL

33 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 33 of 65 The shield will be secured to the TSA using the same studs that attach the TDA s to the composite shell. A hat type fitting along with a threaded nut will be used to secure the micrometeoroid shield/thermal blanket to the attachment stud Mechanical Analysis A NASTRANTM Finite Element Method (FEM) mathematical structural model was developed for the Gamma-ray Large Area Space Telescope (GLAST), Anti-Coincidence Detector (ACD). The FEM, which contains 8,528 elements and 8,899 nodes, was used to determine stress margins, displacements under inertial loads and to obtain the natural modal frequencies of the system. The Tile Shell Assembly (TSA) was modeled using CQUAD4 plate elements that encompassed the honeycomb paneling of the structure. The base frame that supports the TSA was modeled using a combination of CQUAD4 plate elements and CHEXA solid elements. The CHEXA solid elements are within the corner fittings of the base frame. The flexures were modeled using CBAR elements. The overall dimensions of the ACD and the base frame are illustrated in Figure 25. The ACD shell is made from composite honeycomb panels consisting of M46J/RS-3 isotropic laminate facesheets with a Korex (Dupont) 3/ core. The base frame is aluminum 6061-T6, and the flexures are Titanium. The actual tiles that mount onto the ACD were not physically modeled but were represented by non-structural mass. The material properties for the various materials within the FEM are presented in Table 6. Figure 25. Finite Element Model

34 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 34 of 65 Table 6. Material Properties of GLAST components Component Material Young's Modulus (N/m^2) ACD composite facesheets M46J/RS-3 Quasi-Isotropic Laminate 8.96E+10 Shear Modulus (N/m^2) Density (kg/m^3) 1.66E+03 ACD composite core Korex (Dupont), 3/ E E+01 Base frame Aluminum 6061-T6 6.83E E+03 Flexures Titanium Ti-6Al-4V 1.10E E E+03 The total modeled mass of the ACD is kg, including mass contingency for analysis purposes. The mass shown for the TSA includes the tiles and MS/TB, which were represented with non-structural mass. As shown in Table 7, the total mass includes the mass of the electrical cables that were represented with non-structural mass at the sixteen cable tie-down locations. The actual PMT s were not modeled but were also represented with non-structural mass. The mass of the PMT s is included within the mass of the FREE circuit card assembly. The mass breakdown of each component is presented in Table 7. Table 7. Mass Breakdown of ACD Components Component FEM (kg) TSA Flexures 4.42 Base Frame FREE Cables Total ANALYSIS To validate the FEM, a free-free dynamics analysis with a stiffness equilibrium check was performed on the model. This check verified that the model behaves as a rigid body when it is unconstrained. This verification entails obtaining six rigid body modes, which the model did produce when the boundary conditions were removed. In addition to verifying rigid body motion, it also checked the stiffness matrix to verify that it did not contain any grounding effects Static Analysis Static analysis was performed to verify that the system is structurally adequate when subjected to the design limit loads. The components were analyzed using the General Environmental Verification Specification s (GEVS) factors of safety, 1.4 for ultimate and 1.25 for yield. The ACD design limit loads are given in Table 8 and are defined from preliminary in-house GLAST Coupled Loads Analysis specified in the ACD Design and Test Loads Specification, ACD-SPEC The Table 8. ACD Design Limit Loads (G's) ACD Design Limit Loads (G's) Event Direction Liftoff/Transonic MECO (Max Lateral) (Max Axial) Thrust +3.25/ Lateral ± 4.0 ± 0.1 thrust and lateral loads are applied simultaneously for each event in all combinations. There are two events, Liftoff/Transonic (Max Lateral) and MECO (Max Axial). The thrust direction of the loads will be applied along the z-axis and the lateral loads will be applied along the x-y plane (See Figure 25).

35 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 35 of 65 The boundary conditions of the FEM consist of three nodes constrained in 3 DOF to represent the bolt pattern at the corner of the base frame. This connection represents the attachment to the LAT (See Figure 26). The base frame is also constrained at the four mid-side locations with two nodes to represent the two bolts that connect to the grid (See Figure 27). The mid-side nodes are constrained in 3 DOF to represent the connection to the Grid. A detailed stress analysis was performed on the Base Electronics Assembly (BEA). The design limit loads were applied to a combined LAT and ACD model. Two configurations were considered to obtain an envelope of loads for the components to be analyzed. The first configuration assumed the mid-spans (Figure 27) were connected in shear and tension. This case maximized the loads at the mid-span locations. The second configuration assumed that the mid-span connections take tension only, which resulted in maximizing the loads on the corner fittings (Figure 26). Figure 26. Corner Boundary Condition Figure 27. Mid-Side Boundary Condition Once the loads from the FEM were obtained, hand calculations were performed to determine the structural integrity of the corner fittings, the mid-span connections are currently being analyzed. The analysis of the corner fitting showed positive margins (See Table 9) even when conservative assumptions were used. In addition to the stress analysis of the corner-fittings and the mid-span sections, stress analysis was also performed on the tile shell assembly (TSA). Design limit stresses are based on a supported weight including a contingency of at least A fitting factor of 1.15 is included where applicable. A summary of the margins of safety for the TSA Edge clips and the tile flexures are presented in Table 10. Table 9. Margins of Safety for Corner Fitting ITEM FAILURE MODE MS Corner Fitting Bottom Bolts bolt failure 4.4 Corner Fitting Bottom Bolts insert failure 4.3 Corner Fitting ultimate axial loading 0.16 Corner Fitting yield axial loading 0.05 Corner Fitting ultimate shear loading 1.9 Corner Fitting end pad bending 1.3 Corner Fitting Side Bolts bolt failure 9.2 Corner Fitting Side Bolts shear tear out 19.3 Mid-Span Connection to Grid rib failure 0.02 Mid-Span Connection to Grid bolt failure in work

36 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 36 of 65 Detailed stress analysis of the edge clip and concepts for the panel interfaces are in Table 10. The details of the preliminary analysis performed on the flexures are also included in Table 10. Table 10. Margin of Safety of Edge Clips and Flexures PART FAILURE MODE LIMIT STRESS OR FORCE ALLOWABLE Delamination 430 psi 2500 psi Edge Clips ultimate First Ply Failure 1600 psi 28,000 psi ultimate Doublers First Ply Failure 4675 psi 28,000 psi ultimate Bending 20,400 psi 32,000 psi Thermally Induced ultimate Tile Flexure Tension+Bending 1680 psi 32,000 psi Blade ultimate Shear 1470 psi 10,000 psi ultimate Column Buckling 22 lbf 39.5 lbf ultimate Honeycomb Core 153 psi 260 psi Tile Flexure-TSA Crushing ultimate Bonded Interface Adhesive Shear 44 psi 5000 psi yield MS Large Large Large The predicted ACD shell deflections under the design limit loads are tabulated in Table 11. Table 11. Deflections under Design Loads CASE LOCATION DISP (mm) Max Lateral (RSS X & Y) Max Vertical (Z) ACD Shell Side Panel Dynamic Analysis The dynamic analysis was performed to obtain the natural frequency of the system. It was a goal to have the minimum frequency above 50 Hz. The modal frequencies and effective masses are provided in Table 12. The minimum frequency requirement of 50 Hz was achieved. As shown in Table 12, the first natural mode of the system is Hz, which corresponds to the lateral mode of the Tile Shell Assembly (TSA) along the x-axis (See Figure 29). The second mode of Hz is complimentary of the first mode and is along the y-axis. The third mode of Hz is a drum-head mode of the TSA (See Figure 28) ACD Top Panel 0.79

37 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 37 of 65 Table 12. Modal Frequencies and Effective Masses MOMENTS OF INERTIA ABOUT MPFPNT BASIC COORDINATE SYSTEM I-XX I-YY I-ZZ I-XY I-YZ I-ZX 1.43E E E E E E-05 EFFECTIVE MODAL WEIGHTS Frequency X-WT Y-WT Z-WT I-XX I-YY I-ZZ ODE # (Hz) (Kg) (Kg) (Kg) (N-m^2) (N-m^2) (N-m^2) Description E E E-13 Lateral Mode of TSA, X-dir E E E-13 Lateral Mode of TSA, Y-dir E E E-08 Drum-head mode of TSA E E E-12 Side panels of TSA, alternating sides E E E+01 Rotation of TSA, about Z-axis E E E-11 Rotation of TSA, about Y-axis E E E-12 Rotation of TSA, about X-axis E E E-07 2nd Drum-head mode TSA, including sides E E E-19 TSA Side panel mode E E E-13 TSA Side panel mode E E E-05 TSA Side panel mode E E E+01 TSA Side panel mode w/ low mass modes E E E-09 Low mass mode within Baseframe E E E+00 Low mass mode within Baseframe E E E-11 2nd Bending Mode of TSA w/ low mass mode TOTALS E E E+02 % OF TOTAL Figure 29. First Mode Shape Figure 28. Second Mode Shape ANALYSIS SUMMARY The various components that were analyzed showed positive margins of safety once the constraint on the mid-span connection was modified. The dynamic analysis illustrated that the GLAST ACD system met the minimum requirement of 50 Hz. The first frequency of the GLAST ACD is 54.17

38 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 38 of 65 Hz, which corresponds to the lateral mode of the TSA. The displacements of the system, in response to the static loads, were also within the requirements previously stated Thermal Design The ACD thermal design is structured about the LAT Grid and the Tracker towers. The design being a passive one utilizes the grid as a heat sink for the electronics. While the other temperature sensitive component, the Tile Detector assembly, is mainly designed around the effectiveness of the thermal shield and insulation. The ACD electronics, the PMT s and the scintillator tiles have been identified as being temperature sensitive. The ACD contains twelve electronics boards that are mounted to the BEA frame. Each board has been estimated to have a power dissipation of 1.5 watts for a total ACD power dissipation of 18 watts. The operating temperature range that has been established for purposes of Systems Level thermal analyses is from 10 C to 40 C while the survival range is from 20 C to 45 C. The ACD PMT s and scintillator tiles have no power dissipation. The operating temperature range for the scintillator tiles is from 50 C to 40 C. The survival range is from 60 C to 45 C. The thermal design strategy to satisfy the ACD electronics board temperatures is to utilize the LAT aluminum grid structure as a heat sink for the electronic boards. The ICD with LAT specifies that the grid will be operating in a temperature range from 10 C to 25 C. Design analyses adds and subtracts 10C from the grid operating range (-20C to 35 C) to assure a conservative design. For the ACD TDA s, the temperatures will be balanced by the heat flow radiated from the LAT with the heat flow through the MLI blanket. The effective ACD emittance will be approximately A thermal vacuum test will be performed to measure the thermal insulation provided by the micrometeoroid shield by itself without MLI blanket layers. Based on the results of this test and the required insulation characteristics of the ACD, an appropriate number of outer blanket layers will be added to achieve required thermal performance. There are two orbit cases that are used to do the analysis and obtain temperature predictions. The first case being the hot case orbit where either the ±Y face is in the sun for the duration of the orbit. The second case being the cold case where either the ±X face is in the sun. In both cases the +Z is anti-nadir pointing. A safehold case has also been analyzed using the cold case orbit. With the point anytime, anywhere orbit of this mission the bounds of these two cases are sufficient in showing the worst case cold and hot environment. Table 13 displays the temperature predictions for the analysis.

39 LAT-TD D2 LAT Anticoincidence Detector Subsystem Preliminary Design Report page 39 of 65 Table 13. Temperature Predictions Description Cold Operating Temperature Hot Operating Temperature Safehold Temperature Operating Temperature Range Grid Boundary Temperature Trackers Boundary Temperature TDA Cold Temperature to 40 C TDA Hot Temperature to 40 C BEA +X Backplane BEA +X Backplane BEA +X Frame Temp. (-Y end) BEA +X Frame Temp. (middle) BEA +X Frame Temp. (+Y end) BEA +Y Backplane BEA +Y Backplane BEA +Y Frame Temp. (+X end) BEA +Y Frame Temp. (middle) BEA +Y Frame Temp. (-X end) Shell Hot Temperature Shell Cold Temperature The temperature predictions for the X and the Y frame and backplane temperatures are not shown because they are approximately the same as the predictions for the +X and +Y respectively. The thermal model consists of a TSS surface model and a SINDA numerical model. The TSS model is used to calculate orbital fluxes as well as view factors to space. The TSS geometry model contains 34 surfaces with 528 active nodes. Input into the model is the geometry, the optical properties, the environmental properties and the orbit definition. TSS is initiated and radiation couplings to space and heat rates are output from the program. This output is used as input into the SINDA numerical model. The SINDA deck is a numerical network of capacitances and conductances. The SINDA deck contains 615 total nodes. With the conductances, capacitances, interfaces, sources and the output from TSS a SINDA run is initiated and temperatures are output. Shown below are images of the TSS geometry. On the left Figure 30 is the LAT exterior covered with MLI, and Figure 31 is the interior with the grid and tracker towers shown. Figure 30. Exterior MLI Figure 31. Towers and Grid